Chapter Thirteen
Campbell spent the rest of Tuesday, all day Wednesday, and most of Thursday going over each and every factor on Ms. Jennings’ list that could cause a false positive reaction on an HIV antibody test, getting the scientific studies on record. He also called a couple other witnesses Thursday afternoon to add to that list. Sarah dutifully sat there listening to every word. By the time court resumed Friday morning, she wasn’t sure, with so many false positives proven to exist, how anyone could say that a Positive test result is a “true” positive. Then she realized that’s exactly what Campbell wanted the jury to be thinking in this case. He’s doing a good job, she decides.
Armand, on the other hand, had been strangely silent. Before Carolyn Jennings left the stand, Armand asked her a few questions in cross-examination, mainly having her reiterate the point that just because these studies had found certain people with cross-reactions, it didn’t mean, for example, that everyone who had a flu shot a month before taking an HIV antibody test would have a false positive result. But Armand didn’t seem to question the fact that all these false positives could occur. Sarah wondered why.
She also wondered about her brother, and whether Greg might have had any of the things that could have caused a false positive result on his own test. Some of these are very common, she thought. I doubt he had herpes or hepatitis, or I would have known. And we don’t see a lot of these diseases in the U.S. anymore, like leprosy or malaria, or even much tuberculosis.
But there were all those vaccinations listed. She didn’t know, for instance, whether Greg might have gotten a routine tetanus shot before taking his HIV test, maybe after falling off his bike. Or a flu shot; which would make sense, since the whole family would normally get one every year. Or perhaps even a hepatitis vaccine; who knows what their family doctor may have recommended once Greg declared he was gay. Or maybe Greg had a cold when he took the test.
Sarah thought about calling her mother to find out. It had been a while since Sarah had mentioned Greg’s name to her parents, and the last conversation had not gone well. They simply refused to consider any other possibility than Greg dying from HIV/AIDS, because the alternatives were too painful to think about. Sarah knew that her mother would be upset if she asked about different shots and vaccines Greg might have had, and she probably wouldn’t remember that much detail from twenty years ago anyway.
So Sarah stopped in the middle of dialing her cell phone. There’s really no point in making the call, she realized. Even if Mother might remember, and even if the answer is ‘Yes, Greg got a flu shot a couple weeks earlier,’ there’s no guarantee that it had caused a false positive test result. And even if it did, what was that going to change today?
Still, it was the mystery of it all; and as Sarah now sits waiting for court to reconvene, what she longs for more than anything are answers – definitive answers. She already knows Greg died from AZT and not HIV, but would she ever know whether he was really HIV-Positive?
Campbell’s voice interrupts her train of thought; his next witness has taken the stand. “Please state your name and occupation.”
“Wilfred Pullman, Ph.D., Distinguished Professor Emeritus in the Department of Statistics at the University of South Carolina, Columbia.”
“Dr. Pullman, in the course of your career, have you had the opportunity to study the statistics regarding the so-called HIV ELISA antibody test?”
“Yes, I have.”
“Why?”
Pullman chuckles. “It actually began as a hobby about ten years ago. In his book, which I read in 1997, Dr. Peter Duesberg pointed out that while the numbers of AIDS patients were skyrocketing in the late 1980’s and early 1990’s, the number of people infected with HIV remained constant, at least according to the Centers for Disease Control and Prevention. That, of course, made no sense statistically, if HIV were the cause of AIDS; and when I saw that graph from his book, it got me intrigued with the whole subject.”
“And why did you focus on the HIV tests?”
“I had a close friend call me one day. He had just been diagnosed HIV-Positive as a result of one of these tests, and he started asking me a lot of questions I couldn’t answer. So I did some pretty extensive research which has continued over the past few years.”
“And what did that research find?”
“I’m not sure where you want me to begin.” Pullman suddenly looks uncomfortable in the witness chair. Campbell acts quickly to set him at ease.
“Let’s take this chronologically. The HIV ELISA test was put into use in 1985, I believe. What was the earliest study that you found about its statistical accuracy?”
“That would be in 1988. In fact, this is probably the most famous study about the ELISA test, done by a research team headed by a Dr. Burke, which created a lot of chaos at the time.” That was all it took for Pullman to regain his composure.
“And what was this study all about, Dr. Pullman?”
“1.2 million applicants for military service had been given an HIV ELISA test; 12,000 of them tested Positive on their first ELISA – almost exactly 1 percent. But of those 12,000 Positive ELISA tests, only 2000 were ultimately shown to be infected with HIV.”
“And what is the significance of this study?”
“There are actually a number of significant statistics involved. First, the Centers for Disease Control and Prevention say that only .4 percent of our U.S. population is infected with HIV. But 1 percent of the people taking these initial ELISA tests were found HIV-Positive. That’s more than double the results we should get, and that doesn’t speak well of the accuracy of the test itself.”
“A previous witness has testified that the ELISA was intentionally made more sensitive to ensure that no HIV-infected blood got through in our blood supply for transfusions. Wouldn’t that account for a higher than normal statistical result?”
“Yes, you would expect a certain amount of error. But 250%? That’s a little much, don’t you think?”
Way too much, Campbell thinks, but doesn’t say. Instead, he asks, “You said there are other statistical problems?”
“Well, out of the 12,000 Positive initial ELISA results, as I said, only 2000 were later confirmed to be infected with HIV. In other words, there were 10,000 false positive test results. That sets the specificity of the initial ELISA test at less than 20%, and that’s really poor. I don’t know of any other antibody test with that low specificity.”
“And just to be clear, where did these 10,000 false positives come from?”
“They must have been cross-reactions with things other than HIV, like some of the factors you’ve heard about in the last couple of days.”
Campbell points to the large easel that was still standing in the corner of the courtroom with the list of more than sixty medical conditions proven to cause false positive HIV test results. “You mean those?”
“Yes.”
Looking first at the jury and then back to the witness, Campbell continues, “Is there anything else statistically significant about the Burke study?”
“I think that we have to also pay attention to the final result, which is that out of the 1.2 million people tested, 2000 turned out to be HIV-Positive. That’s less than .2 percent of the total tested, and that doesn’t match with the .4 percent of the population that the CDC says are HIV-Positive.”
“And what do you conclude from this?”
“Well, first, we know for a fact that the CDC is simply guessing when they say that .4 percent of the population is HIV-Positive. Even they admit that they don’t have actual Positive test results on all these people. They came up with that number based on extrapolation, meaning that they have a certain number of positive test results from a certain group of people, and they expand that to encompass the entire population. But still, it’s just a guess. And when you line up their guess with the actual results of the Burke study, they don’t match. In fact, the CDC says there are more than twice as many HIV-Positives as the Burke study woul
d indicate. So one of two things could be true. One, the CDC could be wrong and we only have half as many HIV-Positives in the U.S. as they claim – which would be bad for them and affect their funding; or two, the Burke study accurately reflects what happens when you use an antibody test – any antibody test, for that matter – on the general population and not just on high-risk groups.”
“You’ll have to explain that last bit a little more.”
“Basically, whenever you run an antibody test on people who are not at risk for having the virus you’re testing for, you will get a much higher number of false positives. That’s a known fact. In this case, Burke was testing a broad cross-section of military applicants – not just high-risk gay men or IV drug users or hemophiliacs, who have always made up 95% of all HIV-Positives in this country. So it would stand to reason that they would find so many false positives.”
“Are you aware, Dr. Pullman, that in May of 2006, the CDC issued recommendations that every single person in the United States should get an HIV ELISA test, and the American Medical Association endorsed that recommendation?”
“I am aware of that, yes.”
“And, statistically, what will that mean?”
“It will mean that we will see a whole lot more false positive test results than normal. In fact, if we take the Burke study and apply it to the 300 million people in this country who the CDC wants to test, we will get 3 million people who test positive on their initial ELISA. But even the CDC believes that only 1.2 million of them will actually be infected with HIV. That means that we’re going to be telling almost 2 million people that they have tested positive on an HIV ELISA test who are actually false positive. And if the CDC’s guess is wrong about the total number of HIV-Positives in this country, and the Burke statistics are correct when the test is given to low-risk groups, the number of false positives could be as high as two-and-a-half million. That, to me, is statistically outrageous.”
“This Burke study was done in 1988, I think you said. Have there been other studies done since then that either confirm or contradict these results?”
“There have been lots of studies done, yes. Some of them contradict the Burke results, some of them confirm them. Frankly, the numbers range all over the place, and some of that is due to the groups being studied and whether or not they belong to a high-risk group, as I just discussed. But in the end, there is a statistical consensus that emerges.”
“Before we get to that, please tell us about a couple of these other studies, if you will.”
“Well, let’s go to one extreme and talk about a Russian study, done by a Dr. Voevodin and published in 1992 in the prestigious scientific magazine called The Lancet. In 1990, more than 20 million HIV ELISA tests were performed in Russia, and they had about the same results as the Burke study: 20,000, or 1 percent, tested Positive. But later it was found that only 112 were confirmed to be infected with HIV, a much lower percentage than Burke. In 1991, again in Russia, there were 30,000 false positives out of approximately 30 million tests, with only 66 confirmed HIV infections. Now we’re talking about more than 99% false positives. It’s also interesting to note that, in line with the testimony you’ve just had from other witnesses, 8,000 of the 30,000 false positives in Russia in 1991 were reported in pregnant women; but only 6 out of those 8,000 were confirmed to be infected with HIV. Pregnancy, it seems – especially multiple prior pregnancies and births – is a very strong candidate for creating false positive test results.”
Campbell hesitates, wondering whether to go off on a tangent at this point. But the information is too important not to bring up, and he’s not sure when he’ll get a better chance. “Dr. Pullman, I don’t want to digress too far, but are you aware of the statistics of HIV infection coming out of Africa?”
“I’m very aware of them, yes. They’re totally bogus. And one of the reasons is this: There are very few Africans actually being tested for HIV. They don’t have to test anyone in Africa, because, by definition, you don’t have to be HIV-Positive in order to have AIDS on that continent. The main group who does get an HIV test in Africa are pregnant women, and women who have just given birth. As I just said, pregnancy can cause very high false positive test results, so you would expect to see an enormous number of false positives when you test this particular group. Plus, tuberculosis, malaria, and leprosy are on that list you were just discussing that also cause false positive HIV test results, and those three diseases are still very prevalent in sub-Saharan Africa. So we would expect an even higher rate of false positives there than almost anywhere else. But someone, and I don’t know exactly who it is, takes these inflated false positive numbers and extrapolates them to the entire African population and then says that millions of Africans are HIV-Positive. It simply doesn’t hold up under scrutiny.”
“You said that pregnancy was one of those factors that could cause a false positive test result. What are the chances that a woman who has had eight children could test false positive?”
“Extremely high, Mr. Campbell.”
That’s enough of that, Campbell decides. “Dr. Pullman, let’s get back to the other studies you were discussing.”
“Oh, yes. Well, let me mention another one of them. Between 1989 and 1995, about 2700 patients who were undergoing orthopedic surgery received an HIV ELISA test before their procedure. Eight of them tested Positive initially. But only four of those eight were later confirmed, leaving a false positive rate of 50%.”
“A few minutes ago you talked about a ‘statistical consensus’ from all these studies…”
“Yes, I believe there is. Although you can find studies that range from 2% false positives all the way to 99% false positives, it appears than around 50% of all initial ELISA test results will be proven wrong on a second or subsequent ELISA test; and that, in the end, only 20% of initial ELISA positive results will be confirmed as being infected with HIV.”
“Can you put those percentages in real numbers that we can understand a little better?”
“Certainly. Out of 1000 positive results on an initial HIV ELISA test, 500 will be negative on a follow-up ELISA, and only 200 will later be confirmed to be HIV-infected.”
“That means 800 out of 1000 will be false positives?”
“Correct.”
“And if we tested everyone in this country?”
“As I said, if we test all 300 million Americans, as the CDC and the AMA want to do, we’ll get around 2 million false positive test results, maybe more.”
Campbell looks directly at the defendant as he asks the next question. “So you’re saying, Dr. Pullman, that there is only a 20% chance that an initial ELISA test result will be correct.”
“That’s what I’m saying, yes, and what the studies confirm.”
“And in the case of this defendant, in your expert opinion, what are the statistical chances of his initial positive HIV ELISA test being accurate?”
“Like everyone else, I’d say about 1 in 5.”
“Would that also apply to the victim, Beth Ann Brooks? Would the chance of her positive ELISA test being accurate also be about 1 in 5?”
“That would be anyone’s statistical chance, Mr. Campbell. The answer is Yes.”
Campbell turns to look now at Armand. “Nothing further right now.”
“Your witness, Mr. Armand,” the judge instructs.
Armand whispers something to his assistant before getting up from his chair. The assistant is busily looking through some paperwork when Armand asks his first question.
“Dr. Pullman, you mentioned a number of times while you were discussing these studies that some of the initial ELISA tests were later ‘confirmed’ – that’s the exact word you used, I believe – ‘confirmed’ to accurately detect HIV infection, is that correct?”
“Yes, that’s correct.”
“Exactly how were these results ‘confirmed’?”
Pullman and Campbell had earlier discussed the possibility of this kind of question, and Pullman was prepared to
answer truthfully, but carefully.
“That depends on which study you are talking about, Mr. Armand. I believe the ones in Russia were confirmed by actual virus culture, for example, which is why they found so few confirmed results. But there are a number of different tests being used these days to confirm HIV infection after a positive ELISA result.”
“Such as….”
“Objection, Your Honor.” Campbell is on his feet in a flash. “Mr. Armand is trying to get this witness to talk specifically about the Western Blot and viral load tests, among others, and once again there was nothing in my direct that would allow that.”
The judge isn’t so sure. “The witness has clearly talked about confirmation tests, Mr. Campbell.”
“Yes, Your Honor, but only in generic terms. I have no objection if Mr. Armand wants to ask generic questions about ELISA test confirmation, but I must insist again that nothing specific be asked about the Western Blot or viral load tests until I have opened the door to those issues in my direct examination of the appropriate witness.”
The judge raises his eyebrows. “And when will that be, Mr. Campbell?”
“Actually, Your Honor, we’ll be presenting our expert witness on the Western Blot test on Monday.”
Satisfied, “All right. Mr. Armand, you only have the weekend to wait. Do you have other questions of this witness today without specifically talking about these other tests?”
Armand looks at his assistant, who hands him a paper he had found. “Yes, Your Honor.” He then briefly reads the paper, lays it down on his table, and turns back to the witness. “Dr. Pullman, are you aware of the protocol issued by the Centers for Disease Control and Prevention about these ELISA tests?”
“In what regard, Mr. Armand?”
“For example, what does the CDC say about doing multiple ELISA tests?”
“As far as I know, they recommend that a Positive test result on an initial ELISA test be followed up with a second, and then possibly a third ELISA.”
“Are these second and third ELISA tests part of the confirmation procedures you were talking about in these studies?”
“In some of them, yes.”
“And why do you think the CDC suggests that a second or third ELISA test be done?”
“I would assume because of the very poor statistical results of a single ELISA test, Mr. Armand.”
Campbell can’t believe that Armand is taking this approach, but he’s not concerned. Pullman can handle this, he decides.
“Isn’t the reason the CDC wants more than one ELISA test, Dr. Pullman, is to weed out the false positives that might occur in a single test?”
“I can’t speak for the CDC, Mr. Armand, but that would make sense, yes.”
“So, I ask you, Dr. Pullman, as long as these multiple ELISA tests are run to ensure the very minimum of false positive results, what’s the point of all your statistics? I mean, as long as we find those 200 people who are actually infected with HIV in the end, and help them ward off the deadly disease of AIDS, who cares if 800 others were found to have false positive results the first time?”
Campbell can’t believe it. Neither can Pullman.
“Who cares? Perhaps, Mr. Armand, there wouldn’t be a problem if it were as simple as you make it sound, and if AIDS were not such a lethal disease. But in reality, we’re scaring the bejesus out of those 800 people by telling them they might be infected with HIV, and liable to get AIDS and die, when in fact they aren’t. That’s the problem, and I think all of us should care about that.”
Armand immediately picks up the same paper from the table and waves it at the witness. “You must not be fully aware of the CDC protocol, then, Dr. Pullman, because they clearly state that no one should be told they are HIV Positive until their test results have been confirmed.”
“I know that’s what they say, Mr. Armand, but that’s not what actually happens in the real world. People are being told they are HIV-Positive after just a single ELISA test, regardless of what the CDC says, and you know it.” Pullman is getting pissed. Not necessarily a good sign, Campbell thinks. I hope he can keep his cool.
“I don’t know that to be a fact, Dr. Pullman. What I do know is that the CDC clearly says to run confirmation tests before anyone is told they are HIV-Positive. If there is a doctor or two out there who are not following the CDC’s direction, then it is the doctor who is at fault, not the CDC.”
Pullman pulls himself together. “Mr. Armand, in order to run a confirmation test, you have to take more blood from the patient, do you not?” When Armand doesn’t answer right away, Pullman continues. “Just think about it. You know you’ve given blood initially for an HIV test. The doctor or nurse or clinic then calls you and says they need more blood to run more tests. What else are you going to think other than you might have HIV, even if they don’t tell you outright that your test was Positive? What else is going to go through your mind other than you might get AIDS and die soon? What else, Mr. Armand? Tell me!”
Armand quickly says, “No more questions,” and heads back to his table before Pullman can continue. Campbell, on the other hand, is eager to re-direct.
“Dr. Pullman, you seem to be familiar with the CDC protocol Mr. Armand mentioned. Are you also familiar with the protocol from the World Health Organization about running the second or third HIV ELISA test?”
“Yes, I am. They say to take new blood from every patient for each ELISA test done.”
“In other words, if a person has a positive result on their initial ELISA test, the World Health Organization says to draw new blood to run a second ELISA – and again for a third?”
“Correct.”
“Do you know why they say that?”
“To rule out any contamination that may have occurred during the first ELISA test, either when the blood was drawn or the test was processed in the laboratory. If you simply run a second ELISA on the same blood, without ruling out contamination, you could easily get the same wrong result the second, or third, time. That’s simply logical, don’t you think?”
Campbell, of course, doesn’t answer, but nods his head slightly. “So, if someone has a positive result on the very first ELISA, and if the WHO’s protocol is being followed, they’re going to get a call right away to come down and give more blood for further testing.”
“Yes, they are – which simply confirms what I was saying to Mr. Armand, that at a minimum, we’re telling 800 people out of 1000 that they might be infected with HIV when in fact they aren’t, and causing all kinds of problems for them.”
“What kind of problems?”
Pullman smiles. “During my statistical research, I stumbled across the Los Angeles County Department of Health website which had such an interesting answer to that question that I had to write it down, and I carry it around to show people who ask me that.” He pulls a folded piece of paper from his pocket organizer and asks the judge, “Can I read it to you?”
The judge nods approval.
“The question, and this was on their Frequently Asked Questions webpage, is: ‘What are the risks of testing for HIV?’ And their answer is: ‘Risk of testing for HIV includes the following: Taking the HIV antibody test is a stressful event, regardless of the results; and disclosures of an antibody test result, or sometimes the disclosure that a person even took the test, may lead to discrimination, denial of health coverage, stigmatization and violence.’ So I would ask you, Mr. Campbell, if you were told that you might – that you just might – be infected with HIV… that you needed to give more blood and wait another few weeks to find out if it’s true or not… don’t you think you would experience tremendous emotional and psychological trauma, and family stress, along with possible social rejection, even for a few days while waiting for the final results? And we’re talking about putting more than 2 million people in this country through that very thing – all because the HIV ELISA test is so statistically bad. Does that make any sense to you? It doesn’t to me.”
This sounds like a g
ood place to stop, for now, Campbell decides. “Your Honor, I have no more questions at this time. However, I would like to be able to recall this witness at a later date to discuss other statistics that might come up during the course of the trial.”
The judge looks at Armand for an objection, but he only waves his hand in disgust.
“Very well. And since it’s already lunch time on a Friday, this court will stand in recess until ten a.m. Monday morning.”
Are You Positive? Page 17